Predicting uncertain effects of a blast load generated by an accidental vapour cloud explosion

Abstract

Predicting mechanical effects of accidental vapour cloud explosions generated during man-made accidents is of crucial importance to assessing potential damage to property and people endangered by the explosions. Mechanical effects of a vapour cloud explosion are expressed usually in terms of characteristics the blast wave which is generated by explosion and reflected by the exposed structure. The reflected blast wave is described by the pressure signal. The main characteristics of the pressure signal are positive overpressure and positive impulse. These characteristics can be predicted by deterministic mathematical models suggested in the literature. However, these models should be supplemented by quantification of uncertainties related to the positive overpressure and positive impulse. These uncertainties are unavoidable when one deals with such man-made accidents as those involving explosions. The uncertainties can be taken into account by describing inputs of the deterministic models by probability distributions. Uncertainties quantified by these distributions can be propagated to the uncertainties related to the characteristics of the pressure signal. This can be done by applying stochastic simulation (Monte Carlo method). A result of such propagation can be expressed in terms of empirical distributions of the simulated values of positive overpressure and positive impulse.